copyright Peter Stewart, University of Glasgow, 17th May 2017



The data is arranged into subfolders for each figure

All figures were constructed using MATLAB R2013a, but also run in MATLAB R2016a



%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%



/figure2



the static configurations for figure 2 can be calculated using the matlab code `stat1.m' discussed below

the traces shown in figure 2 are provided as .dat files 'tp_rey_R_dl2_3.dat', where R is the Reynolds number

column 1: tension, T
column 2: external pressure, p_e
column 3: length, L_2
column 4: minimum width of the channel, min_x(h^{(s)})
column 5: corresponding x position of the minimum, x_m^{(s)}

the trace for the inviscid limit in figure 2 is provided as .dat files 'tp_rey_inf.dat'

column 1: tension, T
column 2: external pressure, p_e
column 3: minimum width of the channel, min_x(h^{(s)})
column 4: corresponding x position of the minimum, x_m^{(s)}




/figure3



The eigenvalue spectra and corresponding eigenfunctions in figure 3 can be computed using the matlab codes discussed below.

matlab function stat1.m computes the static configuration of the membrane for a given parameter combination, called within the code below

matlab function globfun_final.m computes the eigenvalue spectrum

Input: membrane tension (ten), Reynolds number (rey), external pressure (pe), length of downstream segment (dl2), number of grid points (n), spatial domain (x), grid spacing (dx), initial guess for membrane shape (h0), initial guess for oscillation frequency (beta0)

Output: selected eigenvalues (es), corresponding eigenvectors (vs), corresponding static configuration (h0)




/figure4



The neutral curves can be computed using the codes provided for figure 3




/figure5



/figure6



/figure7



/figure8



/figure9



/figure10